How Does the Broader Autism Phenotype Affect Parent-Mediated Interventions for Children with Autism?
The broader autism phenotype (BAP) describes sub-diagnostic threshold traits of autism/autism spectrum disorder (ASD) that are commonly found in family members of people with autism, but also the general population. BAP is associated with a number of difficulties, including language, face processing, cognitive, social, rigidity and psychiatric conditions. Parent-mediated interventions can benefit children with autism; however, very little research has investigated how the presence of parental BAP might influence parent-mediated intervention and consequently affect outcomes. One small pilot study (n=18; Parr et al., 2015) showed a significant negative correlation between BAP and measures of mother-child interaction post-intervention. The authors concluded that this required further investigation using larger samples, including a control group who received treatment as usual (TAU).
The aim was to determine how parental BAP affects outcomes following parent-mediated intervention or TAU. We hypothesised that BAP would impact parent-child interaction and effectiveness of parent-mediated intervention.
PACT 7-11 was a follow up (FU) of a parent-mediated social communication intervention RCT for children aged 2–4 years with core autism. FU was conducted at a median of 5.75 years from the original trial endpoint (mean age=10.5 years (SD=0.8)). Parental BAP traits, measured at FU using the Family History Interview-Subject (FHI-S) Factor Score, were assumed stable and applied retrospectively for analysis. The following blinded endpoint and FU outcomes were treated as multivariate pairs: Autism Diagnostic Observation Schedule (ADOS) comparative severity scores (CSS); Dyadic Communication Measure for Autism (DCMA) of the proportion parent acts that were synchronous with child attention (DCMA Parent Synchrony); and DCMA proportion of child initiations when interacting with the parent (DCMA Child Initiations). Full Information maximum likelihood structural equation modelling in STATA was used to make use of the full sample and repeat post-randomisation outcomes. Adjusting for baseline outcome value and treatment group, the effect of FHI-S (square root transformed) on each outcome was examined, both as main effect and moderator of treatment effect.
121 (80%) of the 152 trial participants (PACT intervention=59 [77%] of 77, TAU=62 [83%] of 75) completed follow up analysis. Of these, FHI-S scores were available for 102 participants (PACT intervention=49, TAU=53). The models on the full sample confirmed the strong group effects on all outcomes (p<.001) reported in previously published work. FHI-S was not correlated with baseline ADOS CSS, DCMA Parental Synchrony, or DCMA Child Initiations. FHI-S had no effect on outcome ADOS CSS or DCMA Parent Synchrony. There was a marginally significant main effect of FHI-S (p=.079), and a marginal moderation effect (p=.097) on DCMA Child Initiations, and no significant interactions.
In this sample, BAP appears to have a limited role in outcomes following parent-mediated intervention. The finding may be specific to the particular behaviours targeted by the PACT therapy and the relatively low FHI-S scores of the PACT parents. More information is needed to understand how interventions should be tailored to parents, children, and particular environments.